ANA ISABEL DUARTE ANDRÉ
IMPACT OF ENDOCRINE DISRUPTING CHEMICALS IN NUCLEAR
RECEPTOR SIGNALING IN MARINE ORGANISMS:
INVERTEBRATE INSIGHTS
Tese de Candidatura ao grau de Doutor em Ciências do Mar e do Ambiente, Especialidade em Qualidade Ambiental.
Programa Doutoral da Universidade do Porto (Instituto de Ciências Biomédicas Abel Salazar e Faculdade de Ciências), Universidade do Algarve e da Universidade de Aveiro.
Orientador: Doutor Miguel Alberto Fernandes Machado e Santos
Investigador Auxiliar
Centro Interdisciplinar de Investigação Marinha e Ambiental (CIIMAR)
Co-orientador: Doutor Luís Filipe Costa Castro Investigador Auxiliar
Centro Interdisciplinar de Investigação Marinha e Ambiental (CIIMAR)
Esta tese foi financiada por uma bolsa de doutoramento da Fundação para a Ciência e a Tecnologia SFRH/BD/81243/2011 e pelos projetos PTDC/MAR/105199/2008, PTDC/MAR/115199/2009, EXPL/MAR/EST/1540/2012, Norte 2020 e FEDER (Coral - Sustainable Ocean Exploitation - Norte-01-0145-FEDER-000036).
Agradecimentos
A realização desta tese não teria sido possível sem a colaboração e encorajamento de algumas pessoas, às quais gostaria de agradecer:
Começo por agradecer aos meus orientadores, Doutor Miguel Santos e Doutor Filipe Castro, por me terem permitido fazer parte das suas equipas de investigação, pela confiança depositada em mim, por toda a dedicação, ajuda e conhecimentos que me transmitiram ao longo de todos estes anos de trabalho conjunto, assim como pela constante disposição para esclarecimentos, discussões e principalmente pela sua amizade.
Á minha colega Raquel Ruivo pela sua amizade, pelo seu apoio, motivação e ajuda crucial para desenvolver o trabalho.
A todos os membros da equipa, e às pessoas que dela já fizeram parte, do EDEC, AGE e METOX que, de forma direta ou indireta, contribuíram e permitiram a execução deste trabalho, pelo seu companheirismo, amizade e disponibilidade para ajudar sempre que necessário.
À equipa BOGA por ter dado apoio técnico sempre que necessário na execução de ensaios com organismos aquáticos.
À minha família por ter acreditado em mim, por todo o seu apoio, carinho, e por terem possibilitado e ajudado na realização desta tese.
De acordo com o disposto no n.º 1 do artigo 34.º do Decreto-Lei n.º 74/2006, publicado em Diário da República, 1.ª série, n.º 60 de 24 de março de 2006, e republicado pelo Decreto-Lei n.º 115/2013, publicado em Diário da República, 1.ª série, n.º 151 de 7 de agosto de 2013, que procede à terceira alteração ao Decreto-Lei n.º 74/2006, de 24 de março de 2006.
Constam nesta tese os artigos já publicados, que a seguir se discriminam:
Retinoid metabolism in invertebrates: when evolution meets endocrine disruption.
André, A., Ruivo, R., Gesto, M., Castro, L.F., Santos, M.M., 2014. Gen. Comp. Endocrinol., 208, 134-45. DOI: 10.1016/j.ygcen.2014.08.005.
Retinoid level dynamics during gonad recycling in the limpet Patella vulgata.
Gesto, M., Ruivo, R., Páscoa, I., André, A., Castro, L.F., Santos, M.M., 2016. Gen. Comp. Endocrinol., 225, 142-148. DOI: 10.1016/j.ygcen.2015.10.017.
Estes artigos foram adaptados e surgem nesta tese nos Capítulos 1 e 5, respetivamente, respeitando as políticas de direitos de autor da Elsevier. Os autores mantêm o direito de usar conteúdos do artigo (figuras, ilustrações, texto, etc.), no seu todo ou em parte, em teses, sem fins comerciais. Detalhe mais completo sobre as políticas de direitos de autor de Elsevier, pode ser consultado em:
Abstract
The retinoids are lipophilic molecules structural and functional related to vitamin A (or retinol), known to be essential for maintaining critical vertebrate biological processes, such as embryonic development, reproduction, and vision. To maintain such processes vertebrates have a complex mechanism to regulate the spatial and temporal distribution of retinoids that comprises a metabolic (including a synthesis, storage, transport and catabolic cascade) and a signaling pathway. Outside vertebrates, the presence of retinoid metabolic and signaling modules are still poorly understood. Yet, an increasing number of studies have hinted the presence of key molecular players of retinoid metabolic and signaling modules, benefiting from the increasing number of genome sequences from deuterostomes and protostome species. These findings suggest that retinoid pathways might have a more ancient evolutionary origin than previously thought. However, the presence of retinoid-related genes on protostomes, for instance, does not necessarily correlate with a comparable in vivo role as observed in vertebrates. Thus far, studies regarding the identification, isolation, and functional characterization of retinoid metabolic and signaling players have been conducted in a very limited number of protostome species.
In the present work we aimed to improve our knowledge regarding retinoid metabolism and signaling cascade evolution in lophotrochozoan protostomes. Therefore, here we describe the isolation and functional characterization of key molecular metabolic players in marine lophotrochozoan species (annelid and/or mollusks): the β-carotene cleaving oxygenase like enzyme (BCO), the diacylglycerol O-acyltransferase 1 (DGAT1) enzyme and the aldehyde dehydrogenase 1 (ALDH1) enzyme. We showed that the presence of metabolic vertebrate-like key players does not necessarily imply a conserved enzymatic function. 1) The BCO is an enzyme known to be involved in the oxidative cleavage of pro-vitamin A carotenoids into retinoic acid (RA) precursors, retinal (RAL) or β-apocarotenal. We have isolated a BCO-like enzyme from Platynereis dumerilii (annelid). In order to characterize the function of this enzyme, we used an in vivo color shift approach consisting in the expression of the P. dumerilii BCO protein in an Escherichia coli strain, able to produce and accumulating β-carotene. Our results indicate that this enzyme is able to cleave the carotenoids into retinoid precursors. 2) In vertebrates, the DGAT1 enzyme has a role in retinol (ROL) esterification and storage. Here, we isolated a DGAT1 gene orthologue in P. dumerilii and Patella depressa (mollusk) and demonstrated that key amino acid residues involved in vertebrate DGAT1 enzymatic function are fully conserved. In addition, we have determined the Dgat1 gene expression in gonads and digestive gland upon ROL in vivo injection in P. depressa. No changes were observed in Dgat1 gene expression compared to non injected control animals, suggesting that its
major role might not be ROL storage. 3) The ALDH1s are enzymes that irreversible metabolizes RAL into RA. For P. dumerilii and P. depressa we isolated two ALDH1s related enzymes. Through an in vitro protein expression and cell extracts-based biochemical assays we demonstrated that the enzymes are not able to oxidized RAL into RA like their vertebrate’s counterpart. The lack of ability to use RAL as a substrate is in accordance to the key amino acid residues of the substrate entry channel that do not resembles vertebrate’s ALDH1s but other family related proteins.
In parallel, with the isolation of retinoid metabolic modules we also carried out isolation and/or functional characterization of retinoid signaling modules, the nuclear receptors (NR), retinoid X receptors (RXRs) and retinoic acid receptor (RARs) orthologues. The RAR and RXR are transcription factors that mediate the effects of RA isomers by gene transcription regulation. Using an in vitro reporter gene transactivation assay we demonstrated that RXR in P. dumerilii (annelid) and Crassostrea gigas, N. lapillus and Patella vulgata (mollusks) conserve the ability to respond to the putative bone fide ligand 9-cis-RA, in accordance with the conserved amino acid residues of the NR ligand binding domain (LBD). In addition we demonstrated that mollusks RARs are not functionally similar to chordate RAR, since they are not responsive to retinoids. This feature appears conserved within the mollusk phylum. However, the mollusk RAR conserved the ability to form heterodimer with RXR. In this case, mollusks RAR/RXR heterodimer responds in the presence of RAR and RXR ligands by repressing target gene transcription, an action that seems to be ancestral and conserved with mammals.
Overall, our findings suggest that in lophotrochozoans it remains uncertain the exact role of retinoids and their metabolic and signaling modules. Yet, a retinoid metabolic cascade must be present in lophotrochozoans, at least in mollusks, since evidences revealed the presence of several endogenous retinoid precursors and active retinoids. In contrast, the retinoid signaling pathways appear to be conserved throughout evolution among vertebrates and lophotrochozoans taxa, although the functionality of such modules might not be fully conserved.
Although the repertoire of retinoid metabolic and signaling modules are not fully known and characterized in protostomes, evidences supports the conservation of some RA functions between vertebrates and lophotrochozoans. In this work, we also addressed the involvement of the retinoid system in gonad development in the gastropod Patella vulgata. The results are suggestive of an involvement of retinoids in gonad maturation, similar to the role demonstrated in vertebrates.
Environmental endocrine disrupting chemicals (EDCs) are able to interfere with the retinoid pathways. From an ecotoxicological standpoint, most studies have focused on the effects of EDCs mostly on vertebrates with invertebrates receiving less attention. The only
exception is the phenomenon of abnormal development of male reproductive structure in female gastropods, called imposex, that has been associated with organotin compounds (OTs) tributyltin (TBT) and triphenyltin (TPT) exposure. These OTs are known to bind to RXR and modulate their signaling pathways. Few studies have established a link between the retinoid system modulations and reported endocrine negative effects in invertebrate taxa. This aspect is strongly related with the poor knowledge regarding retinoid modules repertoire on invertebrates which consequently hampers our understanding on the wider biological processes and phylogenetic impact of EDCs. Since recent evidences indicate a more ancestral origin of the retinoid system we hypothesize that many more species can also be a target of EDCs. One of our main hypotheses explored in this thesis is related with RXR signaling pathway modulation by TBT and TPT in metazoans. Using an in vitro luciferase reporter assay, we showed that the RXR receptor from the P. dumerilii is also a prime target of TBT and TPT. We also provide evidences that support the hypothesis that in mollusks imposex is mediated by a RXR-dependent signaling cascade. We demonstrated in vitro that TBT, 9-cis-RA, HX630 and methoprene acid, compounds previously described to promote imposex in female N. lapillus, are able to induce target gene transcription activation regulated by RXR. Additionally we have demonstrated that although TPT induce very mild levels of imposex in female N. lapillus at environmentally relevant concentrations, is also a potential endocrine disruptor of RXR signaling pathways in gastropods: It binds to the NR and is able to activate gene transcription, though to a lesser extent than TBT which might explain the different sensitivity among gastropod species.
In parallel we also evaluated if target genes transcription regulated by RAR-dependent signaling pathway would also be a target of modulation by TBT in mollusks. Using an in vitro reported gene transactivation assay we demonstrate that RAR from both mollusks and human, as a monomer and in heterodimer with RXR in the presence of TBT display a gene transcription repression, more significant in the heterodimeric complex. One possible explanation for this response could be related with the TBT ability to bind to RXR and interact with RAR leading to repression in the heterodimer and RAR as monomer. The biological significance of this response remains to be investigated. It might be biological important given that for vertebrates RAR/RXR is implicated in the regulation of numerous physiological processes, and our data indicates that the heterodimeric complex might also be a target of modulation by TBT. Additionally, in the present thesis we also demonstrate that mollusks RAR, as a monomer, loss the ability to response to retinoids and that feature also seems to make then unresponsive to common environmental pollutants (i.e. organochlorine pesticides) that are known to interact with mammalian RARs.
Overall, the present thesis provides novel insights into the evolutionary origin of the retinoid metabolic and signaling pathways as well as into its modulation by EDCs in marine lophotrochozoans.
Resumo
Os retinoides são moléculas lipófilicas, estrutural e funcionalmente relacionados com a vitamina A (ou retinol); reconhecidos por serem essenciais para a manutenção de importantes processos biológicos em vertebrados, tais como o desenvolvimento embrionário, a reprodução e a visão. Para manter esses processos, os vertebrados possuem um mecanismo complexo para a regulação da distribuição espácio-temporal de retinoides que engloba uma via metabólica (incluindo uma cascata de síntese, armazenamento, transporte e catabolismo) e uma via de sinalização. Fora do grupo dos vertebrados, a presença de módulos das vias de metabolismo e de sinalização de retinoides ainda são pouco conhecidas. No entanto, vários estudos têm sugerido a presença de módulos chave do metabolismo e de sinalização de retinoides, beneficiando do crescente número de genomas sequenciados em espécie de deuterostómios e protostómios. Estas evidências sugerem que as vias de sinalização dos retinoides poderão ter uma origem evolutiva mais ancestral do que anteriormente sugerida. No entanto, a presença de genes relacionados com as vias de retinoides em protostómios, por exemplo, não implica necessariamente que tenha uma mesma função in vivo que a reportada em vertebrados. Até à presente data, têm surgido vários estudos focados na identificação, isolamento e caracterização funcional de componentes das vias metabólicas e de sinalização de retinoides mas num número bastante limitado de espécies de invertebrados.
O presente trabalho visa melhorar o nosso conhecimento sobre as vias do metabolismo e da sinalização de retinoides em lofotrocozoários protostómios. No presente trabalho descreve-se o isolamento e a caracterização funcional de alguns dos principais componentes moleculares do metabolismo em espécies lofotrocozoárias marinhas (anelídeos e/ou moluscos): Enzima de clivagem oxidativa de β-caroteno (BCO), a enzima diacilglicerol O-aciltransferase 1 (DGAT1) e enzimas aldeído desidrogenases 1 (ALDH1). Demostrámos que a presença de componentes chave do metabolismo idênticos (ortólogos) a vertebrados não significa necessariamente que apresentem uma função enzimática conservada. 1) A BCO é uma enzima que se encontra envolvida na clivagem oxidativa de pró-vitamina A carotenoides em precursores de ácido retinoico (RA), retinal (RAL) ou β-apocarotenal. Em Platynereis dumerilii (anelídeo), isolamos uma enzima ortóloga da BCO dos vertebrados. De modo a caracterizar esta enzima
funcionalmente, utilizamos uma abordagem in vivo de observação de mudança de cor que consiste na expressão da proteína de BCO de P. dumerilii numa estirpe de Escherichia coli, que apresenta a capacidade de produzir e acumular β-caroteno. Os nosso resultados apontam para que a enzima tenha a capacidade de clivar oxidativamente os carotenoides em percussores de retinoides. 2) Nos vertebrados, a enzima DGAT1 desempenha um papel na esterificação e armazenamento de retinol (ROL). Aqui, isolamos um gene ortólogo de DGAT1 em P. dumerilii e em Patella depressa (molusco). Mostramos que os aminoácidos chave envolvido na função enzimática de DGAT1 em vertebrados estão conservados. Adicionalmente, determinamos a expressão génica de Dgat1 nas gonadas e glândula digestiva de P. depressa 48 h após injeção in vivo de ROL. Não foram observadas diferenças na expressão do gene em comparação com animais controlo não injetados, o que sugere que a enzima possa não apresentar um papel a nível da esterificação e armazenamento de ROL. 3) As ALDH1s são enzimas que metabolizam irreversivelmente RAL em RA. Para P. dumerilii e P. depressa isolamos duas enzimas relacionadas com as ALDH1s. Através de um ensaio baseado na expressão in vitro das proteínas em células e ensaios bioquímicos de extração de extratos celulares demonstramos que as enzimas são incapazes de usar RAL como substrato na síntese de RA como o seu homólogo em vertebrados. A incapacidade de usar RAL como substrato está de acordo com a alteração de aminoácidos chave no canal de entrada do substrato que não são idênticos aos das ALDH1s em vertebrados, mas de outras proteínas relacionadas da família.
Em paralelo com o isolamento de módulos metabólicos dos retinoides, procedemos ao isolamento e/ou á caracterização de módulos das vias de sinalização de retinoides, os recetores nucleares (NR): recetores X do ácido retinoico (RXRs) e recetores do ácido retinoico (RARs). Os RARs e RXRs são fatores de transcrição que medeiam os efeitos dos isómeros de RA ao regularem a transcrição de genes. Usando um ensaio in vitro de transativação do gene repórter da luciferase demonstramos que RXR em P. dumerilii, Nucella lapillus, Crassostrea gigas e P. vulgata conservaram a capacidade de responder ao suposto ligando natural, o 9-cis-RA, de acordo com a conservação de aminoácidos chave no domínio de ligação ao ligando (LBD) do recetor. Adicionalmente demonstramos que os RARs em moluscos não são funcionalmente idênticos aos RARs de cordados, uma vez que não respondem de modo similar à presença de retinoides, e que esta característica parece estar conservada ao longo de todo o filo. Por outro lado, o RAR em moluscos conservou a capacidade de formar heterodímero com o RXR. Neste caso, o heterodímero RAR/RXR em moluscos responde à presença de ligandos de RAR e RXR reprimindo a transcrição de genes alvo, uma ação que em parte parece ser ancestral e conservada com os mamíferos.
De um modo geral os nossos dados sugerem que em lofotrocozoários permanece incerto o papel exato dos retinoides e dos seus módulos metabólicos e de sinalização. Contudo, uma cascata metabólica deverá estar presente em lofotrocozoários, pelo menos em moluscos, uma vez que evidências revelam a presença de vários retinoides endógenos precursores e ativos. Em contraste, as vias de sinalização de retinoides parecem estar conservadas ao longo da evolução entre taxa de vertebrados e lofotrocozoários, embora a funcionalidade desses módulos possa não estar inteiramente conservada.
Embora o repertório dos componentes metabólicos e de sinalização de retinoides não seja totalmente conhecido nem tenha sido inteiramente caracterizado em protostómios, evidências suportam a conservação de algumas funções de RA entre vertebrados e lofotrocozoários. Neste trabalho também foi abordado o envolvimento das vias de sinalização de retinoides no desenvolvimento da gonada no gastrópode Patella vulgata. Os resultados sugerem o envolvimento dos retinoides na maturação da gonada, idêntica à verificada em vertebrados.
Os disruptores endócrinos químicos ambientais (EDCs) possuem a capacidade de interferir com as vias de retinoides. Do ponto de vista ecotoxicológico, muitos estudos têm-se focado no efeito de EDCs sobretudo em vertebrados, recebendo os invertebrados menos atenção. A única exceção é o fenómeno do desenvolvimento anormal de estruturas reprodutoras masculinas em fêmeas de gastrópodes, designado de imposex. Este fenómeno tem sido associado com a exposição a compostos organoestanhos (OTs): o tributilestanho (TBT) e o trifenilestanho (TPT). Estes OTs têm a capacidade de se ligar e de modular as vias de sinalização dependentes de RXR. Um número reduzido de estudos tem estabelecido uma ligação entre a modulação da cascata de retinoides e efeitos endócrinos reportados em taxa de invertebrados. Este facto está em grande parte relacionado com a limitação do conhecimento sobre o repertório de componentes chave das vias de retinoides em invertebrados que consequentemente condicionam a nossa compreensão do impacto de EDCs em processos biológicos. Uma vez que evidências recentes apontam para uma origem evolutiva mais ancestral das vias de síntese e sinalização de retinoides, nós colocamos a hipótese que muitas mais espécies poderão também ser alvo por um mesmo EDCs. Uma das principais hipóteses analisadas nesta tese está relacionada com a modelação do RXR por TBT e TPT em metazoários. Usando um ensaio in vitro de transativação do gene repórter da luciferase, demonstrámos que o RXR de P. dumerilii é também alvo de modulação por TBT e TPT. Apresentámos também evidências que suportam a hipótese de que em moluscos o imposex é mediado por uma cascata de sinalização dependente de RXR. Demonstrando in vitro que o TBT, 9-cis-RA, HX630 e ácido methoprene, compostos anteriormente descritos por apresentarem a
capacidade de promover o desenvolvimento de imposex em fêmeas de N. lapillus, são capazes de induzir a ativação da transcrição de genes alvo mediada por RXR. Adicionalmente também demonstra-mos que embora o TPT não induza o desenvolvimento de imposex em fêmeas de N. lapillus a concentrações ambientalmente relevantes, é também um potencial EDCs das vias de sinalização de RXR em gastrópodes: liga-se e tem a capacidade de ativar a transcrição de genes alvo mas a níveis mais baixos que o TBT, explicando a diferente sensibilidade a este composto entre espécies de gastrópodes.
Em paralelo também avaliámos se a transcrição de genes alvo regulada por via de sinalização dependente de RAR seriam também alvo de modelação por TBT em moluscos. Utilizando o ensaio in vitro de transactivação da transcrição do gene repórter da luciferase mostramos que RAR de moluscos e de humano, tanto em monómero como em heterodímero com RXR na presença de TBT produz uma resposta de repressão da transcrição, sendo esta mais significativa no caso do complexo heterodimérico. Uma possível explicação para esta resposta pode estar relacionada com a capacidade do TBT para se ligar a RXR e interagir com RAR levando a repressão no heterodímero em RAR como monómero. O significado biológico desta resposta permanece por investigar. Esta resposta será biologicamente bastante relevante, uma vez que por exemplo, pelo menos para os vertebrados, RAR/RXR está implicada na regulação de inúmeros processos fisiológicos e os nossos dados prevêm que o complexo heterodimérico possa ser também um alvo da modulação por TBT. Adicionalmente, na presente tese também mostrámos in vitro que RAR de moluscos, enquanto monómero perdeu a capacidade de resposta a retinoides e que essa mesma característica parece também ter levado à perda de resposta a poluentes ambientais comuns (ex. pesticidas organoclorados) que são conhecidos por interagir com RARs em mamíferos.
No geral, a presente tese fornece novos conhecimentos sobre a origem evolutiva das vias metabólicas e de sinalização de retinoides, bem como da sua modulação por EDCs em lofotrocozoários marinhos.
Abbreviations List
ACAT Acyl-CoA: cholesterol acyltransferase
ADHs Alcohol dehydrogenases
AF Activation function
ALDHs Aldehyde dehydrogenases
AP Anterio - posterior
ARAT Acyl-CoA: retinol acyltransferase βc β-carotene
BCO β-carotene oxygenase
BCO-I β-carotene 15,15’-monooxygenase
BCO-II β-carotene 9’,10’-dioxygenase
BLAST Basic Local Alignment Search Tool
BSA Bovine serum albumin
CD2665 4- [6-[(2-Methoxyethoxy) methoxy] -7-tricyclo [3.3.1.13, 7] dec-1-yl-2-naphthalenyl) benzoic acid
CRABP Cellular RA binding proteins
CRALBP Cellular retinaldehyde-binding protein
CRBP-I Cellular retinol-binding protein type I
CRBP-II Cellular retinol-binding protein type II
CNS Central nervous system
CYP26 Cytochrome P450 family 26
DBD DNA binding domain
DGAT1 Diacylglycerol O-acyltransferase 1
DMF N,N-Dimethylformamide
DMSO Dimethyl sulfoxide
DR Direct repeat
DTT DL-Dithiothreitol
EcR Ecdysone receptor
EDCs Endocrine disrupting chemicals
EDTA Ethylenediaminetetraacetic acid
EF-1α elongation factor alpha gene
GSP Gene specific primers
HAT Histone acetyltransferase
HDAC Histone deacetylase activity
HNF4 Hepatocyte nuclear factor 4 receptor
HPLC High-performance liquid chromatography
HPLC-MS High-performance liquid chromatography - mass spectrometry
HSCs Hepatic stellate cells
HX630 4- (7, 8, 9, 10 -Tetrahydro-7, 7, 10, 10-tetramethylbenzo[b] naphtha [2, 3- f] [1, 4]) thiazepin-12-yl-benzoic acid
Ip Isoelectric point
kDa Kilodaltons
LBD Ligand binding domain
LBP Ligand-binding pocket
LC/MS/MS High-performance liquid chromatography -tandem mass spectrometry LE 135 4- (7, 8, 9, 10-Tetrahydro-5, 7, 7, 10, 10 -pentamethyl- 5H-benzo
[e]naphtha [2, 3-b] [1, 4]diazepin-13-yl) benzoic acid
LRAT Lecithin: retinol acyltransferase
LXR Liver X receptor
MA Methoprene acid
MBOAT Membrane-bound O-acyltransferases
MDRs Medium-chain dehydrogenase/reductases family
Mw Molecular weight
NAD+ Nicotinamide adenine dinucleotide
NADP+ Nicotinamide adenine dinucleotide phosphate
NR Nuclear receptor
NR1 Nuclear receptor 1
NR2 Nuclear receptor 2
ORF Open reading frame
OTs Organotin compunds
PCR Polymerase chain reaction
PPAR Peroxisome proliferator-activated receptor
PPARγ Peroxisome proliferator-activated receptor gamma
p,p′, DDE 1, 1-dichloro-2, 2-bis(p-chlorophenyl)ethane
RA Retinoic acid
RACE Rapid amplification of cDNA ends
RAL Retinaldehyde or retinal
RALDH Retinal dehydrogenase
RARE Retinoic acid responsive elements
RBP Retinol binding protein
RDH Retinol dehydrogenase
RE Retinyl ester
REH Retinyl ester hydrolase
ROL Retinol
RP Retinyl palmitate
RT-PCR Real time - polymerase chain reaction
RXR Retinoic X receptor
SDR Short-chain dehydrogenase/reductase
SEC Substrate entry channel
STRA6 Stimulated by retinoic acid protein 6 receptor
TBT Tributyltin
TCDD 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin
TFA Trifluoroacetic acid
ThR Thyroid hormone receptor
TPT Triphenyltin
TTNPB 4 [(E)2 (5, 6, 7, 8Tetrahydro5, 5, 8, 8tetramethyl 2naphthalenyl) -1-propenyl] benzoic acid
TTR Transthyretin
USP Ultraspiracle
INDEX
CHAPTER 1 ... 1
1 General Introduction ... 3
1.1 Retinoids ... 3
1.2 Overview of vertebrate’s retinoid metabolic and signaling pathways ... 5
1.2.1 Retinoid uptake, storage and mobilization pathway ... 5
1.2.2 Cellular uptake and active retinoid synthesis pathways: The canonical route ... 6
1.2.3 RA alternative synthesis pathway: The non-canonical route ... 7
1.2.4 Retinoids signaling pathway ... 8
1.2.5 Retinoids catabolism pathway ... 11
1.3 A comparative analysis of retinoid metabolism in invertebrates: molecular and biochemical evidences ... 11 1.3.1 Tunicates ... 16 1.3.2 Cephalochordates ... 17 1.3.3 Mollusks ... 18 1.3.4 Ecdysozoans ... 20 1.3.5 Porifera ... 23 1.3.6 Other groups ... 23
1.4 Retinoid metabolism: mobilization and storage mechanism evolution ... 25
1.5 Evolution of RA synthesis pathway ... 26
1.6 Evolution of signaling pathway ... 27
1.7 Disruption of retinoid pathawys in Invertebrates ... 28
1.8 Model species ... 32
1.8.1 Nucella lapillus ... 33
1.8.2 Patella vulgata and Patella depressa ... 33
1.8.3 Crassostrea gigas ... 34 1.8.4 Acanthochitona crinita ... 34 1.8.5 Platynereis dumerilii ... 35 1.9 Objectives ... 36 1.10 References ... 38 CHAPTER 2 ... 53
2 Isolation and preliminary functional characterization of a BCO-like gene orthologue in the marine annelid Platynereis dumerilii ... 55
2.1 Abstract ... 55
2.2 Introduction ... 55
2.3 Material and Methods ... 58
2.3.1 Chemical compounds ... 58
2.3.2 BCO isolation ... 58
2.3.3 Phylogenetic analysis ... 61
2.3.5 Retinal HPLC detection ... 63 2.4 Results and discussion ... 63 2.4.1 Isolation of a BCO-like orthologue in P. dumerilii ... 64 2.4.2 In vivo β-carotene cleavage in producing and accumulating E. coli strains and
retinoid detection ... 68 2.5 Conclusion ... 72 2.6 Acknowledgements ... 72 2.7 References ... 72
CHAPTER 3 ... 77
3 Diacylglycerol O-acyltransferase 1 (DGAT1) in lophotrochozoans: insights into the mechanisms of retinol esterification and storage ... 79 3.1 Abstract ... 79 3.2 Introduction ... 79 3.3 Material and methods ... 81 3.3.1 Chemicals ... 81 3.3.2 Animals handling and RNA tissue extraction ... 81 3.3.3 DGAT1 isolation ... 82 3.3.4 Sequence analysis ... 85 3.3.5 Structural organization and transmembrane domain prediction ... 86 3.3.6 Animal handle and experimental exposure ... 86 3.3.7 Tissue RNA extraction ... 86 3.3.8 DGAT1 expression by real-time PCR ... 87 3.3.9 Statistical analysis ... 88 3.4 Results ... 88 3.4.1 Cloning and phylogenetic analysis ... 88 3.4.2 Expression of DGAT1 in P. depressa injected with ROL ... 95 3.5 Discussion ... 96 3.6 Conclusion ... 98 3.7 Acknowledgements ... 98 3.8 References ... 98
CHAPTER 4 ... 103
4 Aldehyde dehydrogenase type 1 enzyme in lophotrochozoans and their implication in retinoic acid synthesis: an evolutionary perspective ... 105 4.1 Abstract ... 105 4.2 Introduction ... 105 4.3 Material and Methods ... 107 4.3.1 Compounds ... 107 4.3.2 RNA extraction and cDNA synthesis in Platynereis dumerilii and Patella
depressa ... 108 4.3.3 Gene isolation in P. dumerilii and P. depressa ... 108
4.3.4 Phylogenetic analysis ... 112 4.3.5 Construction of expression vectors ... 113 4.3.6 Cell culture conditions ... 114 4.3.7 In vitro ALDH activity assays ... 114 4.3.8 Retinoids handle ... 115 4.3.9 Retinoid extraction and analysis ... 115 4.4 Results ... 116 4.4.1 ALDH1- and ALDH2-like gene isolation ... 116 4.4.2 In vitro ALDH1 activity assays ... 129 4.5 Discussion ... 131 4.6 Conclusion ... 135 4.7 Acknowledgements ... 136 4.8 References ... 136
CHAPTER 5 ... 141
5 Retinoid level dynamics during gonad recycling in the limpet Patella vulgata ... 143 5.1 Abstract ... 143 5.2 Introduction ... 143 5.3 Materials and methods ... 145 5.3.1 Animal sampling ... 145 5.3.2 Analysis of retinoid content ... 146 5.3.3 Tissue RNA extraction... 146 5.3.4 pvRAR and pvRXR isolation ... 146 5.3.5 Phylogenetic analysis ... 147 5.3.6 Real-time PCR assays ... 148 5.4 Results ... 148 5.4.1 Retinoid content in P. vulgata gonad ... 148 5.4.2 Cloning and phylogenetic analysis of pvRAR and pvRXR ... 150 5.4.3 Expression of pvRAR and pvRXR ... 152 5.5 Discussion ... 153 5.6 Conclusion ... 157 5.7 Acknowledgments ... 157 5.8 References ... 157
CHAPTER 6 ... 161
6 The Retinoic acid receptor (RAR) in mollusks: function, evolution and endocrine disruption insights ... 163 6.1 Abstract ... 163 6.2 Introduction ... 163 6.3 Material and methods ... 168 6.3.1 Compounds ... 168 6.3.2 Selected mollusk species ... 169
6.3.3 RNA extraction ... 170 6.3.4 RAR isolation in A. crinita ... 170 6.3.5 Phylogenetic analysis ... 171 6.3.6 Construction of plasmid vectors ... 172 6.3.7 Cell culture conditions ... 173 6.3.8 COS-1 cells transfections for in vitro luciferase reporter gene transactivation
assays ... 173 6.3.9 Dual Luciferase Assays ... 174 6.3.10 Statistical analysis ... 174 6.4 Results and discussion ... 175 6.4.1 Cloning a RAR orthologue in the polyplacophora A. crinita ... 176 6.4.2 Phylogenetic and sequence analysis ... 177 6.4.3 Transcriptional activity of mollusks RAR expressed in mammalian COS-1 cells
lines in the presence of retinoid and rexinoids ... 184 6.4.4 Transcriptional activity of mollusks RAR/RXR heterodimer expressed in
mammalian COS-1 cells lines in the presence of rexinoids ... 187 6.4.5 Transcriptional activity of mollusks RAR/RXR heterodimer with NlRAR double
mutation for gain of function, and HsRXRα and NlRXRa loss of ability to bind TBT ... 194 6.4.6 Transcriptional activity of N. lapillus RAR in the presence of common EDCs ... 198 6.5 Conclusion ... 200 6.6 Acknowledgments ... 201 6.7 References ... 201
CHAPTER 7 ... 209
7 Cloning and functional characterization of a retinoid X receptor orthologue in Platynereis dumerilii: insights on evolution and toxicology ... 211 7.1 Abstract ... 211 7.2 Introduction ... 211 7.3 Material and Methods ... 214 7.3.1 Test compounds ... 214 7.3.2 P. dumerilii RXR gene isolation ... 215 7.3.3 P. dumerilii RXR basal tissue expression ... 218 7.3.4 Phylogenetic analysis ... 218 7.3.5 Construction of plasmid vectors ... 219 7.3.6 Cell culture conditions ... 220 7.3.7 COS-1 Cells Transfections for in vitro luciferase reporter gene transactivation
assays ... 220 7.3.8 Dual Luciferase Assays ... 221 7.3.9 Statistical analysis ... 221 7.4 Results ... 222
7.4.1 Cloning and phylogenetic analysis of P. dumerilii RXR ... 222 7.4.2 P. dumerilii RXR tissue expression patterns ... 226 7.4.3 Transactivation assay... 226 7.5 Discussion ... 228 7.6 Conclusion ... 234 7.7 Acknowledgments ... 234 7.8 References ... 234 CHAPTER 8 ... 243
8 Nucella lapillus retinoid X receptors isoforms activate transcription of reporter genes in response to imposex-induction compounds ... 245 8.1 Abstract ... 245 8.2 Introduction ... 245 8.3 Material and methods ... 248 8.3.1 Chemical compounds ... 248 8.3.2 Sequence analysis ... 248 8.3.3 Fusion protein construction ... 248 8.3.4 Cell cultures ... 249 8.3.5 Transativation assays... 249 8.3.6 Ligand binding assay... 250 8.3.7 Statistical analysis ... 251 8.4 Results ... 251 8.4.1 Sequences alignment analysis ... 251 8.4.2 Transcriptional activity of NlRXRa and b isoforms expressed in mammalian
COS-1 cells lines ... 253 8.4.3 TPT Ligand binding assay ... 257 8.5 Discussion ... 258 8.6 Conclusion ... 263 8.7 Acknowledgements ... 263 8.8 References ... 263 CHAPTER 9 ... 269 9 General discussion ... 271 9.1 Background ... 271 9.2 Retinoid metabolic cascade evolution in lophotrochozoans ... 271 9.3 Retinoid homeostasis and reproduction in metazoans: vertebrates and
invertebrates insights ... 274 9.4 Retinoid X receptors and retinoic acid receptors isolation and functional
characterization in lophotrochozoans... 276 9.5 Mechanisms of endocrine disruption of retinoid cascade in lophotrochozoans ... 279 9.6 Conclusion and future perspectives ... 284 9.7 References ... 286
1 General Introduction
1.1 Retinoids
The retinoids are a family of chemical compounds that encompasses molecules structurally and/or functionally related to retinol (ROL) or vitamin A (Blomhoff and Blomhoff, 2006). Structurally, all retinoids encompass a polar end group and a polyunsaturated polyene hydrophobic side chain attached to a cyclic six-carbon ring (Kam et al., 2012). The major natural retinoids occurring, i.e., vitamin A derivatives, are retinyl esters (REs), retinaldehyde (retinal or RAL) and retinoic acid (RA) (Fig. 1.1) (Blomhoff and Blomhoff, 2006). Retinol and REs are the alcohol and esters retinoid forms, respectively. In vertebrates, the first is the predominant retinoid in circulation during the fasting state, serving as a precursor for RAL and RA synthesis; while the former is the main storage form in tissue, also found in the circulation incorporated in lipoprotein particles such as chylomicrons and their remnants (Theodosiou et al., 2010; O’Byrne and Blaner, 2013). Retinyl esters has no biological activity, but it can be converted to active retinoids when needed to satisfy the vitamin A requirements, also serving as substrate for the formation of the visual chromophore 11-cis-retinal (Theodosiou et al., 2010; O’Byrne and Blaner, 2013). Retinal is the aldehyde form; it is a precursor for RA synthesis, but has also a crucial role in vertebrate’s vision, mainly in the metabolism of the pigment rhodopsin (Blomhoff and Blomhoff, 2006; D’Ambrosio et al., 2011). It acts as a hormone controlling the expression of many genes, RA is vitamin A major biologically active form (Blomhoff and Blomhoff, 2006).
Figure 1.1: Chemical structural formulas of naturally occurring retinoids and β -carotene (adapted from
O’Byrne and Blaner, 2013).
In vertebrates, several physiological processes are regulated by retinoids: cell differentiation and apoptosis, embryonic development, growth, reproduction and vision (Clagett-Dame and DeLuca, 2002; Blomhoff and Blomhoff, 2006; D’Ambrosio et al., 2011). Retinoid’s biological importance is well known, being recognized that both the excess and deprivation can have several deleterious effects on vertebrate’s health (Maden et al., 1998; Zile, 1998, 2001; Blomhoff and Blomhoff, 2006). During embryonic development, retinoid imbalance produces congenital malformations, affecting the heart, ocular tissues and several major organ systems: the circulatory, urogenital and respiratory systems as well as the central nervous system (CNS) (Maden et al., 1998; Zile, 1998, 2001). Therefore, vertebrates require an accurate homeostatic control of retinoid levels. This is achieved through a complex network of metabolic and signaling pathways, with ROL playing a pivotal role linking both routes (Cañestro et al., 2006; Theodosiou et al., 2010) (Fig. 1.2). Depending on the homeostatic requirements, ROL can either enter a two-step oxidation cascade, sequentially producing RAL and RA, the biological active metabolite, or undergo esterification in the form of REs to promote retinoid storage; an additional oxidation inactivates RA (Theodosiou et al., 2010). In general, animals are not able to endogenously produce de novo vitamin A, relying exclusively upon ROL, REs, and/or pro-vitamin A (mainly β-carotene) dietary supply (Blomhoff and Blomhoff, 2006).
1.2 Overview of vertebrate’s retinoid metabolic and signaling pathways
1.2.1 Retinoid uptake, storage and mobilization pathway
In vertebrates, retinoid levels are maintained at a homeostatic state through an effective mechanism that provides an adequate vitamin A supply regardless of daily nutritional fluctuation involving a ROL storage and mobilization cascade system (Fig. 1.2) (Blomhoff and Blomhoff, 2006; Theodosiou et al., 2010; Schreiber et al., 2012). Metabolism of vitamin A begins in the intestinal lumen where dietary performed REs are directly hydrolyzed to ROL by the action of retinyl ester hydrolases (REHs) (Schreiber et al., 2012). The resulting ROL is taken up by enterocytes where it is bound to cellular retinol binding protein type II (CRBP-II), re-esterified by lecithin: retinol acyltransferase (LRAT), and incorporated into chylomicrons, along with other dietary lipids and secreted to the general circulation via the lymphatic system (Noy et al., 2000; Blomhoff and Blomhoff, 2006; D’Ambrosio et al., 2011). Circulating REs are transported to storage tissues, such as the liver, white adipose tissue, intestine, lung, kidney and others (D’Ambrosio et al., 2011; O’Byrne and Blaner, 2013). The liver is the main storage organ for ROL. In the hepatocytes, REs are hydrolyzed back to ROL, bound to cellular retinol-binding protein type I (CRBP-I) and delivered to hepatic stellate cells (HSCs) for LRAT-catalyzed esterification and storage (D’Ambrosio et al., 2011; Schreiber et al., 2012). Whereas LRAT is the sole enzyme that esterifies hepatic ROL, in other tissues diacylglycerol O-acyltransferase 1 (DGAT1) and/or other yet unidentified enzymes with acyl-CoA: retinol acyltransferase (ARAT) activity can contribute to ROL esterification (D’Ambrosio et al., 2011). In times of insufficient dietary intake of vitamin A, REs stores are mobilized by REHs into ROL and released into the circulation to supply peripheral tissues (Theodosiou et al., 2010; Schreiber et al., 2012). Before entering circulation, ROL binds to retinol binding protein (RBP) (Blomhoff and Blomhoff, 2006; D’Ambrosio et al., 2011). Generally, ROL-RBP circulates in the blood as a 1:1 molar complex associated with another serum protein, the transthyretin (TTR) to avoid excretion by the kidney (Blomhoff and Blomhoff, 2006; D’Ambrosio et al., 2011).
1.2.2 Cellular uptake and active retinoid synthesis pathways: The canonical route Once in target tissues, a membrane receptor, stimulated by retinoic acid protein 6 receptor (STRA6), mediates ROL cellular uptake (Theodosiou et al., 2010; D’Ambrosio et al., 2011). After delivery to target cells, ROL can be metabolized into RA by two consecutive enzymatic reactions: first ROL is reversibly oxidized producing RAL which is then irreversibly oxidized into RA (Parés et al., 2008; Fig. 1.2). Two enzyme families have been suggested to mediate the first conversion: the cytosolic alcohol dehydrogenases (ADHs), belonging to the medium-chain dehydrogenase/reductases family (MDRs), and the microsomal retinol dehydrogenases (RDHs) included in the short-chain dehydrogenases/reductases family (SDRs) (Duester et al., 2003; Sandell et al., 2007, 2012). Vertebrates present multiple ADH isoforms, but only three are suggested to have a major implication on ROL oxidation process: ADH1, ADH3 and ADH4 (Parés et al., 2008; Theodosiou et al., 2010). In addition to ADHs, RDHs also have a role in ROL oxidation to RAL, mainly the RDH10 (Sandell et al., 2007; Farjo et al., 2011; Sandell et al., 2012). The second step in RA synthesis involves the RAL irreversible oxidation into RA (Fig. 1.2). This reaction is carried out by ALDHs. There are generally three ALDHs class in vertebrates, which are named as ALDH1a1, ALDH1a2 and ALDH1a3 (Parés et al., 2008). For long, the specificity of RA synthesis was thought to reside exclusively at the level of the second reaction step with ALDH1a being considered the sole rate limiting enzyme; However, RDH10-mediated oxidation of ROL also plays an important role in the control and regulation of RA production as does the subsequent ALDH1a-mediated reaction establishing a novel nodal point in RA feedback regulation (Sandell et al., 2007; Farjo et al., 2011; Sandell et al., 2012).
Figure 1.2: Main components of retinoid metabolic and signaling cascade in vertebrates (adapted from
André et al., 2014).
1.2.3 RA alternative synthesis pathway: The non-canonical route
Some animals are also able to satisfy their own vitamin A requirement by cleaving pro-vitamin A carotenoids derived from fruits and vegetables diet source (Kiefer et al., 2001; Simões-Costa et al., 2008; Amengual et al., 2011). The β-carotene (βc) is the most abundant carotenoid present in the diet and tissues, and is one of the main retinoids synthesis precursors (von Lintig, 2010). In order to generate retinoids, βc must be cleaved. Two major enzymes, the β-carotene 15, 15’- monooxygenase (BCO-I) and β -carotene 9’, 10’-dioxygenase (BCO-II), controls vitamin A synthesis from dietary βc and other carotenoids (von Lintig and Vogt, 2000; Kiefer et al., 2001). The BCO-I oxidative cleaves βc symmetrically at the 15, 15' carbon double bond yielding two molecules of RAL, which can then be either oxidized to RA or/and reduced to ROL by the enzymes from the canonical route (von Lintig, 2010). In contrast, BCO-II cleaved βc asymmetrically generating a molecule of β-ionone ring and one of β-apocarotenal (Kiefer et al., 2001). The β-apocarotenal is then converted to β-apocarotenoic acid that is stepwise oxidized to RA in a process involving enzymes that have yet to be identified (Kiefer et al., 2001).
1.2.4 Retinoids signaling pathway
The resulting RA from βc asymmetric cleavage and ALDH1as activity can have two possible fates: being inactivated by degradation or used for biological functions regulation. Thus, cytosol RA levels, like ROL, are regulated by specific cellular binding proteins, the cellular retinoic acid binding proteins (CRABPs) type I and II. CRABP-I delivers RA for degradation, whereas CRABP-II transports RA to the cell nucleus, mediating their binding to nuclear receptor (NR) superfamily members (Blomhoff and Blomhoff, 2006). Retinoic acid serves as ligand to the retinoic acid receptors (RARs) and/or retinoid X receptors (RXRs), regulating the transcription of several genes (Szanto et al., 2004; Blomhoff and Blomhoff, 2006). For each retinoid receptor, vertebrates have three distinct isoforms: RXRα (NR2B1), RXRβ (NR2B2), RXRγ (NR2B3), RARα (NR1B1), RARβ (NR1B2), and RARγ (NR1B3) (Theodosiou et al., 2010). Both all-trans-RA and 9-cis-RA isomers serve as ligands for all RARs, whereas for RXRs only 9-9-cis-RA specifically bind with high affinity (Mangelsdorf et al., 1992; Allenby et al., 1993; Germain et al., 2006a, b).
Like other members of NRs superfamily, RARs and RXRs are organized in five to six modular regions: A/B, C, D, E and/or F (Aranda and Pascual, 2001; Germain et al., 2006a) (Fig. 1.3). The A/B region is a poorly conserved N-terminal domain that contains a constitutively active transactivation function 1 (AF-1) to which the coactivators bind (Germain et al., 2003). The C region is highly conserved, and it corresponds to the DNA-binding domain (DBD); its small motifs, the P-, D- and T-box are important for conferring sequence-specific DNA recognition and ligation for the response elements in the promotor region of a target gene and/or dimerization (Germain et al., 2003; Bastien and Rochette-Egly, 2004; Germain et al., 2006a). The C region is connected to the E region via a flexible hinge, the so-called D domain (Bastien and Rochette-Egly, 2004; Germain et al., 2006a, b). The D domain is poorly conserved, and it allows C and E domains to adopt different conformations (Bastien and Rochette-Egly, 2004; Germain et al., 2006). The E region is also known as the ligand binding domain (LBD) and mediates transcription activation of target genes (Aranda and Pascual, 2001; Germain et al., 2006a). The LBD harbors four structurally surfaces that has distinct function: 1) a dimerization surface, which interacts with LBDs from other NRs; 2) a surface that recognize and interacts with ligands, termed as ligand-binding pocket (LBP); 3) a surface that interacts and binds to cofactors (coactivators or corepressors) to regulate target gene transcription and 4) an activation function, so-called AF-2, which mediates ligand-dependent transactivation (Germain et al., 2003; Germain et al., 2006a). RARs also has an F domain, which is absent in RXRs, the function still remains unknown (Aranda and Pascual, 2001; Bastien and Rochette-Egly, 2004; Dawson and Xia, 2012).
Figure 1.3: Schematic modular structure of nuclear receptors. A/B domain, with an AF-1 activation
function located in the N-terminal. A highly conserved DNA-binding domain, the C domain. D domain, a highly flexible hinge region. E domain responsible for ligand-binding and converting NRs to active forms that bind to DNA, which holds a AF-2 activation domain. A F domain with an unknown function (adapted from Aranda and Pascual, 2001).
To modulate RAR-dependent gene transcription, RXRs serve as an obligate heterodimeric partners to RARs. The RXR is required for efficient binding of RAR/RXR heterodimer to the retinoic acid response elements (RAREs) in the regulatory regions of target genes (Fig. 1.4) (Blomhoff and Blomhoff, 2006; Germain et al., 2006b). The heterodimer RAR/RXR RAREs consists in two direct repeats of the nucleotide sequence, (A/G)G(G/T)TCA separated by a variable number of nucleotides (usually, one, two or five) (Bastien and Rochette-Egly, 2004; Szanto et al., 2004; McGrane, 2007). In the absence of RA, RAR/RXR heterodimer is bounded to RAREs and to a corepressor complex including a histone deacetylase activity (HDAC)/SIN3 complex, recruited via corepressors NCoR or SMRT, resulting in chromatin condensation and gene transcription silencing (Fig. 1.4) (Bastien and Rochette-Egly, 2004). When RA enters the cell nucleus, it binds to RARs and/or RXR, then a conformational change occurs. The corepressors dissociate and a coactivator complex possessing histone acetyltransferase (HAT), methyltransferase, kinase or ATP-dependent remodeling activities (SWI/SNF). These activities lead to chromatin decondensation and core transcription factors recruitment; facilitating the positioning of the transcriptional machinery at the promoter region, that ultimately leads to gene transcription initiation (Fig. 1.4) (Bastien and Rochette-Egly, 2004).
Figure 1.4: Mechanism of target gene regulation by RAR/RXR heterodimers. A) In the absence of ligand,
the heterodimer RAR/RXR is associated with RARE recruiting corepressors complexes having an enzymatic activity of deacetylation of histones, which results in chromatin condensation and transcription repression. B) After RA binding to RAR, the heterodimer suffers a conformational change, releasing the corepressor complex and recruiting a coactivator complex with an enzymatic transacetilase activity of histones, resulting in chromatin descondensation and thus allowing the activation of target genes transcription (adapted from Marlétaz et al., 2006).
Apart from forming a heterodimer complex with RARs, RXRs also have the ability to functions as a dimer with itself (homodimer - RXR/RXR) or act in heterodimeric complexes with other NRs members, such as peroxisome proliferator-activated receptors (PPAR), thyroid hormone receptor (ThR), liver X receptor (LXR) (Szanto et al., 2004; Germain et al., 2006b). Therefore, RXRs are implicated in many signaling pathways apart from those involved in retinoid signaling, regulating transcription of several genes, and concomitantly being involved in a wide range of physiological functions, such as homeostasis, development and metabolism (Szanto et al., 2004). In the case of RXRs transcription regulation as a homodimer complex (RXR/RXR), the signaling pathways are believed to be 9-cis-RA binding dependent, but its exact function remains uncertain (Mangelsdorf et al., 1992; Allenby et al., 1993; Germain et al., 2006b).
In addition to RA functions mediated through NRs signaling pathway, their action via non-genomic mechanisms has also been reported, such as activation of protein kinase cascades, which influence gene expression through phosphorylation processes (Blomhoff and Blomhoff, 2006; Al Tanoury et al., 2013).
1.2.5 Retinoids catabolism pathway
Given its robust role in gene transcription regulation, RA catabolic inactivation is required to maintain a proper homeostatic equilibrium. This oxidative step is mainly catalyzed by members of the cytochrome P450 family 26 (CYP26) (Thatcher et al., 2010). The CYP26 enzymes family receives RA that is believed to be bound to cellular retinoic acid binding proteins (CRABPs), promoting the catabolism of RA by metabolizing into polar, and less potent, metabolites, including 4-hydroxy retinoic acid, 4-oxo retinoic acid, 18-hydroxy, 5, 6-epoxy RA or 5, 8-epoxy RA, and others (Blomhoff and Blomhoff, 2006; Theodosiou et al., 2010). The biological relevance of these metabolites as signaling molecules is still unclear, but there are reports that might have a biological activity (Thatcher et al., 2010; Theodosiou et al., 2010).
1.3 A comparative analysis of retinoid metabolism in invertebrates: molecular and biochemical evidences
In comparison with vertebrates, the information regarding the presence of an elaborated retinoid system remains fragmented and incomplete for invertebrate lineages. Initially, most of the retinoid key molecular components were described outside chordates. Therefore, retinoid metabolic and signaling pathways were considered to be a chordate innovation, probably related to the evolutionary origin of their body plan organization (Fujiwara and Kawamura, 2003; Simões-Costa et al., 2008). However, evidences from massive genome sequencing effort, regarding phylogenetically informative species, allowed to identify retinoid-related genes, from the metabolic and signaling cascade, and question this traditional view (Fig. 1.5) (e.g. Cañestro et al., 2006; Albalat and Cañestro, 2009; Theodosiou et al., 2010). In addition, although conducted in a limited number of invertebrate taxa, the endogenous retinoid profile has also been investigated, again supporting the assumption of a retinoid cascade presence outside vertebrates (Table 1.1). Altogether, findings point to an early emergence in basal bilaterians (Cañestro et al., 2006; Albalat and Cañestro, 2009; Theodosiou et al., 2010). Yet, the origin and evolution of retinoid pathways is still poorly understood. Retinoid cascade is not yet entirely characterized in most invertebrate phyla. For instance, the existence of invertebrate genes orthologues does not necessarily correlate with in vivo vertebrate identical roles at the level of retinoid metabolism and/or signaling (Albalat, 2009). Thus, so far studies regarding the identification, isolation, and functional characterization of key molecular modules have been restricted to a few species (Albalat, 2009; Theodosiou et al., 2010; Gesto et al., 2012, 2013). Studies on the physiology of retinoids are also scarce, or even non-existing, for most invertebrate groups. This current lack of knowledge limits our basic understanding regarding the organizational, functional similarities and differences
between invertebrate and vertebrate retinoid systems. Findings clearly point to an older ancestry of retinoid pathways than previously thought, despite some fundamental contrasts. The main focus has been drawn from urochordates, cephalochordates and mollusks. Most importantly, available reports suggest a mosaic pattern when compared to vertebrate’s retinoid functions, with conserved aspects (e.g. REs storage in some mollusk species) and key differences (e.g. RAR presence and signaling in ecdysozoans and gastropods). In order to enhance our insights on retinoid biology and pathway evolution, future studies should focus on the isolation and functional characterization of the related molecular components and on the detection of endogenous retinoids.
In this section, the current knowledge on retinoid pathways in invertebrate phyla is presented.
Table 1.1: Overview of polar and non-polar retinoids as well as β-carotene presence in invertebrates. (+) Present in the species; (−) absent; (na) not analyzed. HPLC – high-performance liquid chromatography; HPLC–MS – high-performance liquid chromatography–mass spectrometry; GS/MS – gas chromatography–mass spectrometry.
Specie Phylum; Class Tissue RAL ROL REs
all-
trans-RA
9-cis-RA
13-cis-RA βc Analytic method Ref.
Polyandrocarpa misakiensis
Chordate; Ascidiacea
Developing bud Adults all body
+ + - - na na + + na na - - na na HPLC Kawamura et al., 1993 Halocynthia roretzi Chordate; Ascidiacea Eggs Gonads Hepatopancreas Gill Hemolymph cells Hemolymph plasma
Body wall muscles + + + + + + + + + + + + + + + + + + + + + na na na na na na na na na na na na na na na na na na na na na na na na na na na na HPLC Irie et al., 2003, 2004 Branchiostoma floridae Chordate;
Leptocardii All body + + na + + + na HPLC Dalfó et al.,
2002
Locusta
migratoria Arthropoda; Insecta; Embryos na na na + + na na
HPLC-MS Nowickyj et al.,
2008
Uca pugilator
Arthropoda;
Malacostraca Blastema of
regenerating limbs + na na + + - na HPLC and GS/MS
Hopkins, 2001; Hopkins et al., 2008 Litopenaeus vannamei Arthropoda; Malacostraca Ovary Digestive gland + + + na na na na na na na na na + + Diode array spectrophotometer and HPLC Paniagua-Michel and Liñan-Cabello, 2000; Liñán-Cabello et al., 2003 Lymnaea stagnalis Mollusca; Gastropoda CNS na na na + + na na HPLC-MS Dmetrichuk et al., 2008
Osilinus lineatus Mollusca; Gastropoda Gonads Digestive gland na na + + + + + + + + + + na na HPLC and HPLC-MS Gesto et al., 2012
Patella depressa Mollusca;
Gastropoda Gonads Digestive gland na na + + + + + + + + + + + na Spectroscopically, HPLC and HPLC-MS Goodwin, 1950; Gesto et al., 2013
Nucella lapillus Mollusca;
Gastropoda Gonads Digestive gland Kidney Gill Prostate CNS Sperm-ingest gland Albumen gland Capsule gland na na na na na na na na na - - - - - - - - - - - - - - - - - - + + na na na + na na na + + na na na + na na na + + na na na + na na na na na na na na na na na na
HPLC and HPLC-MS Gesto et al., 2013 Nassarius reticulatus Mollusca; Gastropoda Complex gonads/
digestive gland na - - + + + na HPLC and HPLC-MS
Gesto et al., 2013
Geodia cydonium
Porifera;
Demospongiae All body na na + + - + + HPLC
Biesalski et al., 1992
Figure 1.5: Cladogram representing the current knowledge concerning the presence/absence of key molecular components in different metazoan groups. (+)
indicates that molecular components are present in the phylum/species; (−) means that molecular components are absent from genome for analyzed species; (unk) absence of evidences (a) indicates that the molecular component is present in some species but absence in others from the same phylum. Hypothetic phylogenetic relationships according to Evans and Gundersen-Rindal (2003) and Theodosiou et al. (2010).
1.3.1 Tunicates
Tunicates are vertebrates’ closest living relatives. However, the few available studies conducted so far suggest that retinoid spatiotemporal distribution is differently controlled in these two animal groups (Irie et al., 2003, 2004).
In Ciona intestinalis the presence of putative vertebrate-like components involved in RA synthesis, including ROL and RAL oxidation, NR-mediated signaling and RA catabolism were genome predicted, whereas a retinoid storage system seems to be lacking (Theodosiou et al., 2010). However, the majority of these genetic modules were not isolated or functionally characterized. Among the few cloned genes is a β-carotene oxygenase (BCO) orthologue, identified in the visual system of larvae and adult C. intestinalis (Nakashima et al., 2003; Takimoto et al., 2006). Escherichia coli purified BCO-like exhibited significant asymmetrical carotenoid cleavage activity, since no RAL was detected (Poliakov et al., 2012). A gene coding for a putative cellular retinaldehyde-binding protein like (CRALBP) expressing in the central nervous system (CNS) during embryonic development, was also identified (Nakashima et al., 2003).
In Polyandrocarpa misakiensis both RXR and RAR homologues were isolated (Hisata et al., 1998; Kamimura et al., 2000); whereas for Botrylloides leachi only a RAR homologue was identified (Rinkevich et al., 2007). RXR has also been cloned and its expression evaluated in Halocynthia roretzi (Maeng et al., 2012). Regarding these species, reports suggest a major role for RA as an endogenous signaling molecule, participating in the morphallactic bud development regulation and tissue regeneration (Kawamura et al., 1993; Kamimura et al., 2000; Rinkevich et al., 2007; Kaneko et al., 2010). RAR is, in fact, expressed during B. leachi whole body regeneration; RAR knockdown resulted in regeneration arrest and bud malformation. The same pattern was reported when RA synthesis was inhibited (Rinkevich et al., 2007). The presence of a RAR orthologue, as well as RA synthetic and catabolic enzymes ALDH1 (ALDH1a/b/c/d) and CYP26, has been also reported in C. intestinalis (Nagatomo and Fujiwara, 2003; Theodosiou et al., 2010; Sobreira et al., 2011). Interestingly, independent losses of the RA genetic machinery might have occurred amongst larvaceans, as suggested by the absence of ALDH1a, CYP26 and RAR orthologues in Oikopleura dioica genome (Cañestro et al., 2006). Consistently, treatment with RA, a classical morphogen, did not induce homeotic posteriorization of anterior structures; likewise, no developmental abnormalities were observed after treatment with an ALDH inhibitor (Cañestro and Postlethwait, 2007). Thus, O. dioica develops and maintains an RA-independent anterio– posterior (AP) axial patterning, typical of a chordate body plan (Cañestro and Postlethwait, 2007).
The available data on endogenous retinoid content highlights a major difference in the synthesis, storage and homeostatic mechanisms between ascidians and vertebrates. In contrast to the latter, RAL is seemingly the main storage and transport retinoid in tunicates, being hypothesized that it is rapidly metabolized to RA when required (Irie et al., 2004). In agreement, endogenous levels of all-trans-RA and RAL were detected in developing buds and adult P. misakiensis (Kawamura et al., 1993). Also, in H. roretzi eggs and adult tissues the most abundant retinoid was RAL and practically no ROL or REs could be detected (Irie et al., 2003, 2004). Recently, RA has been reported to participate in body patterning during embryo development in C. intestinalis (Pasini et al., 2012), supporting the biological relevance of active endogenous retinoids in most urochordates.
1.3.2 Cephalochordates
Comparatively, amphioxus is recognized to have a vertebrate-like response to RA, including the associated homeostatic control mechanisms and signaling pathways (Escriva et al., 2002; Marlétaz et al., 2006). Exhaustive genome analyses have identified various RA metabolic and signaling components in Branchiostoma floridae (Albalat et al., 2011; Poliakov et al., 2012). In some cases genes were isolated and their function assessed. For instance, an ADH3 orthologue was cloned and in situ hybridization analysis revealed restricted expression patterns: adult gut, embryos, late free-swimming and feeding larvae (Cañestro et al., 2000). No functional characterization of ADH3 was performed to elucidate its possible role in ROL oxidation. Two retinol dehydrogenase orthologues, RDH1 and RDH2, were identified and functional characterized. Both RDHs exhibited dehydrogenase activity catalyzing the reduction of RAL into all-trans-ROL, using NADH as cofactor (Dalfó et al., 2007).
RAR and RXR orthologues were also isolated and shown to mediate RA-dependent transcription (Escriva et al., 2002; Tocchini-Valentini et al., 2009). The ability of RAR/RXR heterodimers to recognize retinoic acid responsive elements (RARE) was validated using electrophoretic mobility shift assays (Escriva et al., 2002). In the presence of RA the heterodimer activated transcription (Escriva et al., 2002). Using a cell-based assay, 9-cis-RA was reported to activate RXR, albeit at relative higher concentrations (Tocchini-Valentini et al., 2009). B. floridae putatively exhibits pathways for both β -carotene and ROL-derived RA synthesis and also RA catabolism, given that five BCO-like genes, ALDH1 isoforms (ALDH1a/b/c/d/e/f), SDRs and CYP26 orthologues were genome predicted (Albalat and Cañestro, 2009; Theodosiou et al., 2010; Albalat et al., 2011; Sobreira et al., 2011). Mechanisms for REs hydrolysis and ROL handling also seems to be present, since putative REHs and CRPB were genome predicted (Theodosiou et al., 2010). Regarding the counter esterification pathway, DGAT1 orthologues, but not LRAT,
can be retrieved in B. floridae genome (Albalat et al., 2011; Poliakov et al., 2012). B. floridae also seems to lack putative, STRA6, TTR and RBP orthologues, suggesting alternative transport and cellular uptake mechanisms (Theodosiou et al., 2010; Albalat et al., 2011).
In the adult B. floridae, endogenous levels of ROL, RAL and RA isomers were reported and appear to be involved in morphogenetic processes (Dalfó et al., 2002; Escriva et al., 2002; Schubert et al., 2004). The RA synthesis pathway is also active (Dalfó et al., 2002). In fact, ROL-exposed amphioxus developed morphological abnormalities as observed in RA-exposed animals (Dalfó et al., 2002). Further data suggests an important role for RA signaling in amphioxus nervous system AP patterning and development of epidermal sensory neurons. During embryonic development RAR is expressed in the CNS and strongly upregulated upon RA treatment; conversely, total down-regulation of RAR was reported in the presence of specific antagonists (Escriva et al., 2002). Regarding AP positioning of epidermal sensory neurons, treatments with RA and RAR antagonists affected the collinear expression of Hox genes, up- and down regulating expression, respectively (Schubert et al., 2004).
1.3.3 Mollusks
An increasing body of knowledge advocates for the presence of putative retinoid metabolic and signaling pathways in mollusks. However, studies have focused mainly in NR identification, cloning and their implication in a very limited number of biological functions. Although the detection of endogenous retinoids predicted that, in mollusks, retinoid pathways are active and present, the isolation of the associated genetic machinery and its functional characterization is still missing.
Genome searches have identified retinoid metabolic modules in the herbivore Lottia gigantea (Albalat and Cañestro, 2009; Theodosiou et al., 2010; Sobreira et al., 2011). This species hypothetically exhibits mechanisms for both β-carotene and ROL-derived RA synthesis and RA catabolism as BCO, ADHs, SDHs, three ALDH1a (ALDH1a/b/c) and CYP26 orthologues were genome predicted (Albalat and Cañestro, 2009; Theodosiou et al., 2010; Sobreira et al., 2011). Yet, L. gigantea appears to lack vertebrate-like retinoid storage, transport and mobilization system (Theodosiou et al., 2010). It remains to be investigated whether the referred enzymes are functionally similar to the vertebrate orthologues. For instance, in the dogwhelk Nucella lapillus, ADH3 was isolated and, unlike the ubiquitous vertebrate orthologue, has expression mostly in the digestive gland (Coelho et al., 2012).
The study of retinoid signaling pathways in mollusks emerged with the identification and cloning of RXR orthologues in diverse gastropod species, such as the
